Forum EPW http://epwforum.uk/ 

electronphonon matrix elements http://epwforum.uk/viewtopic.php?f=7&t=731 
Page 1 of 1 
Author:  Fairoja [ Wed Nov 21, 2018 8:56 am ] 
Post subject:  electronphonon matrix elements 
Dear All, I am a participant of 'School on ElectronPhonon Physics from First Principles' held at the International Centre for Theoretical Physics in Trieste on 1923 March 2018. It was a very good school and I had a great experience attending the School. On Tuesdays Handson section, we tried to compute the electronphonon matrix elements using a slightly modified phonon program (which includes printing information). It seems like the output from those calculations are 'g(qpoint, frequency, band1, band2)'. The gmatrix should be a function of kpoint also, which is missing in the output. I found on the subroutine 'elphel' it is mentioned that  write(6,*) write(6,*) 'We here write only the matrix elements for k=Gamma in the first proc:' write(6,*) ! ! consider only initial state k=0  Does this mean that all the g written on the output file are for k=0? In this case how one can get g as s function of kpoint as well. Thank you for your time and consideration, Best regards, Fairoja.  Fairoja Cheenicode Kabeer Postdoctoral researcher, Materials Theory, Uppsala University. 
Author:  sponce [ Thu Nov 22, 2018 10:25 am ] 
Post subject:  Re: electronphonon matrix elements 
Dear Fairoja, You can consider any state. We only show the elph matrix element at k=\Gamma and then changed q. However you can look at any kpoint. In the code you have: Code: ! consider only initial state k=0 ik = 1 WRITE(6,'(a,3f10.6)') 'xk ', (xk(n,ik),n=1,3) where ik=1 is in this case the Gamma point. You can just change that value to get different kpoints. Note that you can also add kpoint loop around all of that like this Code: DO ik=1, MAX_NUMBER_OF_K_POINTS WRITE(6,'(a,3f10.6)') 'xk ', (xk(n,ik),n=1,3) write(6,*) ' ibnd jbnd imode eig_i (eV) eig_j (eV) omega_nu (meV) g (meV)' ... ENDDO Be careful that this will generate very large txt files (lots of data). Best wishes, Samuel 
Author:  SaraFiore [ Wed Jan 09, 2019 8:56 am ] 
Post subject:  Re: electronphonon matrix elements 
Dear sponce, I am interested in the same topic. I downloaded the version released for the school in the subroutine './tests/Inputs/t07/elphon.f90' there is Code: ! consider only initial state k=0 ik = 2 WRITE(6,'(a,3f10.6)') 'xk ', (xk(n,ik),n=1,3) ! What does this '2' represent? how can i change it if I am interested in specific k point? Thank you in advance for any help Sara Fiore ETH Zürich  PhD Student  Doctorate at DITET Rechn. Modellierung Nanostrukturen 
Author:  sponce [ Fri Jan 11, 2019 12:26 pm ] 
Post subject:  Re: electronphonon matrix elements 
Dear Sara Fiore, This is the specific kpoint to be printed (ik=2 is the second kpoint (of the local pool, so better to do this with 1 cpu)). The kpoint value (x,y,z coordinate) will be reported in the output with the line WRITE(6,'(a,3f10.6)') 'xk ', (xk(n,ik),n=1,3) Best, Samuel 
Author:  Fairoja [ Wed Jan 16, 2019 11:26 am ] 
Post subject:  Re: electronphonon matrix elements 
Dear Samual, Thank you for the reply! I tried to modify the output writing part of the program (subroutine elphsum2 of elphon.f90) and performed some calculations. I have tried different qpoints q1 0.000000000000000 0.000000000000000 0.000000000000000 q2 0.000000000000000 0.000000000000000 0.119721397504348 q3 0.000000000000000 0.000000000000000 0.239442795008697 q4 0.000000000000000 0.125000000000000 0.125000000000000 ... The first three finished without any issue, but the output file for the q4 has some NaN values written. The following is the style of output  ibnd jbnd imode omega_nu (meV) # loop over kpoints g (meV) eig_i(k) (eV) eig_j(q) (eV) # end loop  54.268873 0.000000 0.000000 55.053020 0.000000 0.000000 56.387030 0.000000 0.000000 NaN 0.000000 NaN 55.010279 0.000000 0.000000 54.644403 0.000000 0.000000 .... 4.733780 12316.262843 53834.825476 NaN 55197.660572 NaN 30.164047 0.000000 0.000000 30.031159 0.000000 0.000000 30.054339 0.000000 0.000000 59.273177*************** 0.000000 27.924379 0.000000 0.000000 Can you please comment on this issue. Is it due to the writing issue of eig_j(q)? if yes, How can I solve this? Thank you very much for your time and consideration, Fairoja.  Fairoja Cheenicode Kabeer Postdoctoral researcher, Uppsala University. 
Page 1 of 1  All times are UTC 
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ 